Initial management of advanced non-small cell lung cancer lacking a driver mutation

INTRODUCTION — Lung cancer is the leading cause of cancer-related mortality in the United States and worldwide. More than 80 percent of lung cancers are classified as non-small cell lung cancer (NSCLC). Although targeted therapies have redefined treatment options for patients with molecularly defined NSCLC (eg, epidermal growth factor receptor-mutant, anaplastic lymphoma kinase-rearranged NSCLC), these therapies are ineffective in those whose tumors lack such genetic alterations, who comprise the majority of NSCLC patients. However, immunotherapy has become integrated into the treatment of such patients, which has led to improvements in survival and quality of life.

This topic reviews the initial approach to management of patients with nondriver-mutated NSCLC. An overview of the approach to advanced NSCLC, discussions of management of driver-mutated NSCLC, management of brain metastases in NSCLC, and subsequent line treatment of NSCLC are all found separately.

●(See "Overview of the initial treatment of advanced non-small cell lung cancer".)

●(See "Personalized, genotype-directed therapy for advanced non-small cell lung cancer".)

●(See "Anaplastic lymphoma kinase (ALK) fusion oncogene positive non-small cell lung cancer".)

●(See "Systemic therapy for advanced non-small cell lung cancer with an activating mutation in the epidermal growth factor receptor".)

●(See "Brain metastases in non-small cell lung cancer".)

●(See "Subsequent line therapy in non-small cell lung cancer lacking a driver mutation".)

FACTORS IN CHOOSING INITIAL THERAPY — Treatments for patients with advanced NSCLC are palliative, and thus intended to prolong survival and preserve quality of life for as long as possible, while minimizing the side effects due to treatment. Immune checkpoint inhibitors targeting either programmed cell death protein 1 or programmed cell death ligand 1 (PD-L1) have become routinely part of the clinical approach for management of NSCLC. Rationale for use and mechanism of action of checkpoint inhibitors in cancer is discussed elsewhere. (See "Principles of cancer immunotherapy".)

Factors that affect choice of treatment in NSCLC that lacks a driver mutation include the level of PD-L1 expression, the extent of disease, and histology. Our approach is consistent with expert guidelines from the American Society of Clinical Oncology and Ontario Health (Cancer Care Ontario) [1].

●Patients with PD-L1 expression ≥50 percent are offered either monotherapy with a checkpoint inhibitor or a platinum-doublet chemotherapy plus a checkpoint inhibitor.

●For patients with PD-L1 expression <50 percent, the combination of a platinum-doublet chemotherapy and a checkpoint inhibitor is standard.

●Choice of chemotherapy, for those receiving it, is influenced by histology. (See 'Agent to partner with platinum' below.)

PD-L1 LOW (<50 PERCENT) OR UNSELECTED TUMORS — Pembrolizumab is used for the frontline treatment of metastatic nonsquamous NSCLC in combination with pemetrexed and carboplatin, irrespective of programmed cell death ligand 1 (PD-L1) expression. This combination is our preferred treatment option for those with PD-L1 expression <50 percent or unknown advanced adenocarcinomas. However, there are multiple appropriate alternatives with regulatory approval, as discussed in the sections below. (See 'Alternatives' below.)

We suggest pembrolizumab in combination with a carboplatin/taxane doublet in patients with advanced squamous cell cancers with low or absent PD-L1 expression, although atezolizumab-based combinations are also available. (See 'In squamous cancers' below.)

Rationale for choice of chemotherapy is found below. (See 'Choice of chemotherapy to pair with checkpoint inhibitor' below.)

Irrespective of histology, nivolumab plus ipilimumab is an appropriate alternative to these regimens, particularly if a chemotherapy-free option is desired. (See 'Nivolumab plus ipilimumab, with or without chemotherapy' below.)

Pembrolizumab plus chemotherapy (preferred)

In nonsquamous cancers — The frontline addition of pembrolizumab (200 mg intravenously every three weeks) to platinum-doublet chemotherapy (platinum agent and pemetrexed) has improved outcomes relative to chemotherapy alone in two randomized trials enrolling patients with advanced, nonsquamous NSCLC.

In the phase III KEYNOTE-189 trial, 616 patients with advanced, PD-L1-unselected, nonsquamous NSCLC were randomized in a 2:1 ratio to chemotherapy (cisplatin or carboplatin with pemetrexed) with or without pembrolizumab [2]. Those receiving chemotherapy alone could receive pembrolizumab upon progression. At a median follow-up of 10.5 months, the addition of pembrolizumab to chemotherapy improved 12-month overall survival (OS) rates relative to chemotherapy alone (69 versus 49 percent; hazard ratio [HR] for death 0.49, 95% CI 0.38-0.64). Twelve-month OS improvements were observed in all PD-L1 categories, with the greatest numerical differences observed in PD-L1-expressing tumors:

●<1 percent – 62 versus 52 percent (HR 0.59, 95% CI 0.38-0.92)

●1 to 49 percent – 72 versus 51 percent (HR 0.55, 95% CI 0.34-0.90)

●≥50 percent – 73 versus 48 percent (HR 0.42, 95% CI 0.26-0.68)

Similar observations were made at longer follow-up [3,4]; five-year OS rates were 19 versus 11 percent [5].

Median progression-free survival (PFS), the other primary endpoint, was also improved with the addition of pembrolizumab (8.8 versus 4.9 months; HR for disease progression or death 0.52, 95% CI 0.43-0.64). The HR for PFS was less than 1 across all subgroups of PD-L1 expression (HR 0.36 for PD-L1 ≥50 percent; HR 0.55 for PD-L1 1 to 49 percent; and HR 0.75 for PD-L1 <1 percent), but the differences were not statistically significant in the group with PD-L1 <1 percent.

Among all patients in the trial, objective response rate (ORR) was improved with the addition of pembrolizumab (48 versus 19 percent). Severe adverse events (≥grade 3) occurred in 67 percent of the patients in the pembrolizumab-combination group and in 66 percent of those in the placebo-combination group, with improved patient-reported quality of life [6].

These results confirmed findings of the earlier phase II KEYNOTE-021 trial [7].

In squamous cancers — The addition of frontline pembrolizumab to platinum-based doublet chemotherapy has also improved outcomes in squamous cell cancers, without adding substantial toxicity. Atezolizumab plus chemotherapy is investigational for squamous NSCLCs. (See 'Atezolizumab plus carboplatin and a taxane (in nonsquamous cancers)' below.)

In the phase III KEYNOTE-407 trial, 559 patients with PD-L1-unselected, treatment-naïve, advanced squamous NSCLC were randomized in a 1:1 ratio to chemotherapy (carboplatin with either paclitaxel or nabpaclitaxel) with either pembrolizumab or placebo. Outcomes were improved with pembrolizumab plus chemotherapy, with five-year OS rates of 18 versus 9.7 percent, respectively (HR 0.71, 95% CI 0.59-0.85) [8].

In earlier reporting, the median OS with and without pembrolizumab, by tumor PD-L1 expression was [9]:

●<1 percent – 15.9 versus 10.2 months (HR 0.61, 95% CI 0.38-0.98)

●1 to 49 percent – 14 versus 11.6 months (HR 0.57, 95% CI 0.36-0.9)

●≥50 percent – Not reached (NR) versus NR; (HR 0.64, 95% CI 0.37-1.10)

Although the OS benefit did not achieve statistical significance among those with the highest level of PD-L1 expression (≥50 percent), analysis was limited by the small number of patients in this subgroup (n = 146).

The median PFS with and without pembrolizumab, by tumor PD-L1 expression was:

●<1 percent – 6.3 versus 5.3 months (HR 0.68, 95% CI 0.47-0.98)

●1 to 49 percent – 7.2 versus 5.2 months (HR 0.56, 95% CI 0.39-0.8)

●>50 percent – 8 versus 4.2 months (HR 0.37, 95% CI 0.24-0.58)

In the overall group, the ORR was 58 percent (95% CI 51.9-63.8) in the pembrolizumab-combination group and 38 percent (95% CI 32.7-44.4) in the placebo-combination group. Severe adverse events (≥grade 3) occurred in 70 percent of the patients in the pembrolizumab-combination group and in 68 percent of those in the placebo-combination group.

Alternatives — Although we prefer pembrolizumab and chemotherapy for most patients with PD-L1-low tumors (<50 percent), multiple other regimens are available that have regulatory approval in the United States and are appropriate alternatives.

Atezolizumab plus carboplatin and a taxane (in nonsquamous cancers) — Atezolizumab is an anti-PD-L1 antibody engineered to avoid antibody-dependent, cell-mediated cytotoxicity of activated T cells that may express PD-L1. Atezolizumab, in combination with either carboplatin, paclitaxel, and bevacizumab (table 1), or atezolizumab with carboplatin and nabpaclitaxel, is an alternative to the pembrolizumab/chemotherapy based combinations in nonsquamous cancers lacking an epidermal growth factor receptor (EGFR) or anaplastic lymphoma kinase (ALK) mutation [10]. (See 'Pembrolizumab plus chemotherapy (preferred)' above.)

●Atezolizumab, carboplatin, paclitaxel, and bevacizumab – The IMpower 150 trial randomly assigned 1202 patients with PD-L1-unselected, advanced, nonsquamous NSCLC to first-line chemotherapy (carboplatin and paclitaxel) combined with either atezolizumab (ACP), atezolizumab plus bevacizumab (ABCP), or bevacizumab (BCP) [11]. Crossover on progression was not allowed. Among 692 EGFR/ALK wild-type patients, those receiving ABCP experienced improved PFS (8.3 versus 6.8 months; HR for disease progression or death 0.62, 95% CI 0.52-0.74) and OS (19.2 versus 14.7 months; HR for death 0.78, 95% CI 0.64-0.96) relative to those receiving BCP. The PFS benefit of ABCP over BCP was also observed among patients with baseline liver metastases [12]. Patients across arms rated common symptoms with chemotherapy and immunotherapy similarly [13]. In KEYNOTE-189, this subgroup also experienced improvements with the combination of pembrolizumab and chemotherapy [2], suggesting that immunotherapy plus chemotherapy is a viable treatment option for this subset.

The most common grade 3 or 4 treatment-related adverse events were neutropenia, febrile neutropenia, and hypertension. Rash, stomatitis, febrile neutropenia, and hemoptysis were higher among those receiving ABCP than among those in the BCP group by <10 percentage points. This regimen has now been US Food and Drug Administration (FDA) approved as a treatment alternative for driver mutation-negative patients with advanced, nonsquamous NSCLC.

●Atezolizumab, carboplatin, and nabpaclitaxel – Other trials have also shown benefits with the addition of atezolizumab to nabpaclitaxel and carboplatin, which is also an acceptable treatment regimen in nonsquamous cancers [10]. In IMpower 130, adding atezolizumab to carboplatin plus nabpaclitaxel in the frontline setting for advanced nonsquamous NSCLC improved both PFS (7.0 versus 5.5 months; HR 0.64) and OS (18.6 versus 13.9 months; HR 0.79), differences that were statistically significant [14]. Although an atypical chemotherapy backbone in the nonsquamous NSCLC setting, nabpaclitaxel was chosen because, unlike paclitaxel, it does not require steroid premedication, which may affect the response to an immune checkpoint inhibitor. This regimen is approved by the FDA.

Separately, in IMpower 132, atezolizumab plus pemetrexed-based chemotherapy improved median PFS in nonsquamous cancers (7.6 versus 5.2 months; HR 0.60, 95% CI 0.49-0.72) relative to chemotherapy alone, with a nonsignificant trend toward improved OS (18.1 versus 13.5 months; HR 0.46, 95% CI 0.64-1.03) [15]. However, given the lack of regulatory approval and statistically significant survival benefit, atezolizumab plus pemetrexed is not one of our favored regimens.

Nivolumab plus ipilimumab, with or without chemotherapy — Although we typically use the pembrolizumab- or atezolizumab-based strategies discussed above, given cross-trial comparisons of toxicity and efficacy, nivolumab plus ipilimumab is another FDA approved option for adult patients with metastatic NSCLC without EGFR or ALK genomic tumor alterations with PD-L1 expression ≥1 percent; nivolumab and ipilimumab is also approved with two cycles of platinum-based chemotherapy in tumors lacking EGFR and ALK alterations, irrespective of PD-L1 expression [16].

●Nivolumab plus ipilimumab – Nivolumab plus ipilimumab has demonstrated a superior survival compared with chemotherapy in the CheckMate-227 trial, irrespective of PD-L1 expression [17]. Among 1739 patients with chemotherapy-naïve NSCLC across the spectrum of tumor PD-L1 expression, the median duration of OS was 17.1 months with nivolumab plus ipilimumab versus 13.9 months with chemotherapy (HR 0.73, 95% CI 0.64-0.84) [17]. Improvements were similar across the spectrum of PD-L1 expression (for those with PD-L1 <1 percent, HR 0.70; PD-L1 ≥ 1 percent, HR 0.79; and PD-L1 ≥50 percent, HR 0.70).

The percentage of patients with grade 3 or 4 treatment-related adverse events in the overall population was 33 percent with nivolumab plus ipilimumab and 36 percent with chemotherapy. Treatment-related adverse events leading to discontinuation were more frequent with nivolumab plus ipilimumab versus chemotherapy (18 versus 9 percent, respectively). The most common immunologic adverse events of any grade among those receiving nivolumab plus ipilimumab were skin reactions (34 percent) and endocrine events (24 percent). Treatment-related deaths occurred in eight patients who received nivolumab plus ipilimumab and in six patients who received chemotherapy.

●Nivolumab plus ipilimumab in combination with chemotherapy – Nivolumab in combination with ipilimumab and platinum-doublet chemotherapy was evaluated in the CheckMate-9LA trial, in which 719 patients were randomly assigned to either two cycles of platinum-doublet chemotherapy plus nivolumab and ipilimumab (until progression), or four cycles of platinum doublet chemotherapy without immunotherapy [16,18]. Patients receiving nivolumab/ipilimumab/chemotherapy experienced improved median OS (15.6 versus 10.9 months; HR 0.66, 95% CI 0.55-0.80), PFS (6.8 versus 5.0 months; HR 0.70, 95% CI 0.57-0.86) and overall response rate (38 versus 25 percent), relative to those receiving chemotherapy alone.

Serious adverse reactions occurred in 57 percent of patients receiving nivolumab, ipilimumab, and chemotherapy, with the most frequent (>2 percent) being pneumonia, diarrhea, febrile neutropenia, anemia, acute kidney injury, musculoskeletal pain, dyspnea, pneumonitis, and respiratory failure. Fatal adverse reactions occurred in seven patients (2 percent) and included hepatic toxicity, acute renal failure, sepsis, pneumonitis, diarrhea, and hemoptysis in the setting of thrombocytopenia.

Cemiplimab plus chemotherapy — Cemiplimab has FDA approval for use in combination with platinum-based chemotherapy for patients with NSCLC lacking an EGFR, ALK, or ROS1 alteration [19], based on preliminary results from a clinical trial demonstrating improved survival relative to chemotherapy alone. It likely has similar outcomes relative to pembrolizumab and other checkpoint inhibitors, when used in combination with chemotherapy, although there have been no direct comparisons.

In preliminary results of the randomized trial Study 16113, among 466 patients assigned to chemotherapy with or without cemiplimab, the cemiplimab group experienced improvements in median OS (22 versus 13 months; HR 0.71, 95% CI 0.53-0.93), PFS (8.2 versus 5.0 months; HR 0.56, 95% CI 0.44-0.70), and ORR (43 versus 23 percent) [20]. Serious adverse events occurred in 25 percent of patients in the cemiplimab group, with the most frequent being pneumonia, anemia, and neutropenia, and 6 percent were fatal.

Tremelimumab and durvalumab plus chemotherapy — Tremelimumab and durvalumab have regulatory approval in combination with platinum-based chemotherapy in patients with metastatic NSCLC with no sensitizing EGFR or ALK genetic alteration [21,22].

In the open-label POSEIDON trial, 1013 patients with EGFR/ALK wild-type tumors were randomly assigned to tremelimumab plus durvalumab plus chemotherapy, durvalumab plus chemotherapy, or chemotherapy alone [23]. The three-drug combination performed better than chemotherapy alone, both in regards to OS (median 14.0 versus 11.7 months; HR 0.77, 95% CI 0.65-0.92) and PFS (median 6.2 versus 4.8 months; HR 0.72, 95% CI 0.60-0.86), with grade 3 to 4 adverse effects occurring in 52 and 44 percent of patients, respectively. Although durvalumab plus chemotherapy (without tremelimumab) improved PFS compared with chemotherapy (HR 0.74), differences in OS were not statistically significant.

Given these results, we consider tremelimumab and durvalumab plus chemotherapy an acceptable frontline option in metastatic NSCLC lacking an EGFR or ALK mutation.

Less favored options — Although we prefer pembrolizumab and chemotherapy for most patients with PD-L1-low tumors (1 to 49 percent), there is also evidence to consider using pembrolizumab monotherapy as an alternative in patients with low PD-L1 expression (but not in PD-L1-negative tumors). We only consider this option for patients with intermediate PD-L1 expression who decline chemotherapy; in such cases, pembrolizumab monotherapy is an acceptable, but likely inferior, option compared with pembrolizumab and chemotherapy.

In the phase III KEYNOTE-042 trial of patients with treatment-naïve, advanced, EGFR/ALK wild-type NSCLC and at least 1 percent tumor PD-L1 expression, pembrolizumab monotherapy was compared with platinum-doublet chemotherapy, chosen according to histology. Of note, this trial did not permit crossover. At a median follow-up of 12.8 months, results are as follows [24]; OS, the primary endpoint, was prolonged among patients receiving pembrolizumab compared with those receiving chemotherapy as follows, according to PD-L1 expression:

●≥50 percent (599 patients), 20 versus 12 months (HR 0.69, 95% CI 0.56-0.85)

●≥20 percent (818 patients), 18 versus 13 months (HR 0.77, 95% CI 0.64-0.92)

●≥1 percent (1274 patients), 17 versus 12 months (HR 0.81, 95% CI 0.71-0.93)

●In exploratory analysis of survival among patients with PD-L1 expression between 1 and 49 percent, OS was 13.4 versus 12.1 months (HR 0.92, 95% CI 0.77-1.11)

Longer follow up confirmed these benefits, with five-year OS rates of 22, 19, and 17 percent with pembrolizumab for tumors with ≥50 percent, ≥20 percent, and ≥1 percent PD-L1 expression [25].

There was no statistically significant PFS benefit among patients receiving pembrolizumab compared with those receiving chemotherapy, except for those with the highest level of PD-L1 expression (≥50 percent; median PFS 7.1 versus 6.4 months [HR 0.81, 95% CI 0.67-0.99]). As in KEYNOTE-024, pembrolizumab was better tolerated than chemotherapy, with grade 3 to 5 adverse events occurring in 18 percent of those receiving pembrolizumab and 41 percent of those receiving chemotherapy.

Overall, the survival benefit with pembrolizumab seen amongst all patients with >1 percent PD-L1 expression seems to be largely driven by the subgroup of patients with high (≥50 percent) PD-L1 expression. Moreover, among those with intermediate PD-L1 expression (1 to 49 percent), chemotherapy resulted in similar OS as pembrolizumab despite the fact that most of the patients assigned to chemotherapy did not receive immunotherapy upon progression, which is considered standard of care.

Choice of chemotherapy to pair with checkpoint inhibitor — For those with a good performance status in whom chemotherapy is indicated, combination chemotherapy regimens using a platinum compound (cisplatin, carboplatin) plus a second active cytotoxic agent are preferred.

Multiple cytotoxic agents in addition to cisplatin and carboplatin have antitumor activity. Our typical approach is as follows.

Preference for platinum-based regimens — Although the improvement with platinum-based chemotherapy is small compared with a nonplatinum-based doublet (table 2), we agree with the recommendation from the American Society of Clinical Oncology guidelines that first-line therapy should consist of a platinum-based doublet regimen [26].

A meta-analysis that included 4792 patients treated in 17 randomized trials found that the use of a platinum-based doublet regimen was associated with a slightly higher survival at one year (relative risk [RR] for survival 1.08, 95% CI 1.01-1.16) [27]. When the trials were analyzed according to the platinum compound used, cisplatin was associated with a significantly higher survival at one year compared with nonplatinum regimens (RR for survival 1.16, 95% CI 1.06-1.27). By contrast, carboplatin-based regimens were not better than nonplatinum regimens (RR for survival 0.95, 95% CI 0.85-1.07).

Carboplatin typically preferred over cisplatin — Carboplatin is often substituted for cisplatin in combination regimens for patients with advanced NSCLC due to its more favorable toxicity profile. Overall, the evidence suggests that cisplatin gives a higher ORR than carboplatin, although the survival benefit is small and may not be clinically meaningful. We believe carboplatin is appropriate for the majority of patients with advanced disease, for whom quality of life is an important consideration. Moreover, for patients with both nonsquamous and squamous advanced NSCLC, carboplatin-based doublets have been studied in combination with immunotherapy, but cisplatin-based doublets have not. (See 'PD-L1 low (<50 percent) or unselected tumors' above.)

In a meta-analysis that included individual patient data from 2968 patients in nine trials, substitution of carboplatin for cisplatin was associated with a lower response rate (24 versus 30 percent; odds ratio 1.37, 95% CI 1.16-1.61) although differences in survival were not statistically significant (median 8.4 versus 9.1 months; HR for death 1.07, 95% CI 0.99-1.15) [28]. Cisplatin-containing regimens were associated with more nephrotoxicity, nausea, and vomiting, while carboplatin combinations caused more severe thrombocytopenia. Improvements in the management of chemotherapy-induced vomiting and the use of lower doses of cisplatin in newer regimens have mitigated some of the concerns about cisplatin toxicity [29]. (See "Prevention of chemotherapy-induced nausea and vomiting in adults".)

Careful adherence to full doses of carboplatin may obviate any possible efficacy differences between carboplatin and cisplatin. In a phase III trial, 1363 patients with advanced NSCLC were randomly assigned to gemcitabine plus one of three platinum regimens: cisplatin (50 mg/m²), cisplatin (80 mg/m²), and carboplatin (area under the curve 6) [30]. Median survivals for the three treatment arms were 8.3, 9.5, and 10.0 months, respectively; the trial met the predetermined criteria for noninferiority of the carboplatin regimen.

Cisplatin remains the preferred platinum compound when systemic therapy is used with curative intent in conjunction with other treatments (ie, adjuvant therapy, combined-modality treatment of stage III disease). (See "Systemic therapy in resectable non-small cell lung cancer" and "Management of stage III non-small cell lung cancer".)

Schedule — Although most regimens use a three- or four-week interval between treatment courses, weekly administration at lower doses can alter the toxicity profile [31,32].

Although further data are needed, a weekly schedule of paclitaxel plus carboplatin may be considered if neurotoxicity is a particular concern.

This approach was examined in a phase III trial in which 883 chemotherapy-naïve patients were randomly assigned to paclitaxel plus carboplatin, administered either every three weeks or on a weekly schedule with reduced doses (table 3) [31]. Severe neurotoxicity was significantly less frequent with the weekly schedule compared with every three weeks (4.4 versus 9.1 percent), but grade 3 or 4 diarrhea was more common (4.2 versus 1.1 percent). The efficacy of the weekly schedule was not significantly different (response rate 38 versus 33 percent; median time to progression 6.1 versus 7.2 months; and median survival 8.9 versus 9.5 months).

Agent to partner with platinum — Earlier data suggest improved outcomes with platinum combinations over a platinum alone.

In a meta-analysis that included 13,601 patients from 65 randomized trials, two-drug regimens significantly increased both response rate and OS compared with single-agent chemotherapy (ORR 26 versus 13 percent with single-agent therapy; one-year survival rate 35 versus 30 percent) [33].

Histology provides insight into the optimal agents to combine with a platinum compound.

Nonsquamous histology — Regimens containing pemetrexed are more effective in patients with adenocarcinoma than regimens lacking pemetrexed. Alternatives to pemetrexed include taxanes and gemcitabine.

The impact of histology on the results with platinum-based chemotherapy was illustrated by a phase III trial in which cisplatin plus pemetrexed was compared with cisplatin plus gemcitabine as initial therapy [34,35]. Survival in the 847 patients with adenocarcinoma was significantly prolonged with cisplatin plus pemetrexed compared with cisplatin plus gemcitabine (median 12.6 versus 10.9 months). Conversely, cisplatin plus gemcitabine was superior to cisplatin plus pemetrexed in the 473 patients with squamous cell carcinoma (median 10.8 versus 9.4 months).

The extent to which these results represent increased activity for pemetrexed in patients with adenocarcinoma or decreased activity of pemetrexed in those with squamous cell carcinoma is unknown. There are insufficient data to compare pemetrexed-based regimens with other combinations in patients with adenocarcinoma.

Squamous histology — Multiple agents pair effectively with platinums. These include taxanes (docetaxel, paclitaxel, nanoparticle albumin-bound paclitaxel [nabpaclitaxel]), gemcitabine, vinorelbine, and camptothecins (irinotecan, topotecan). We do not use pemetrexed in squamous cell cancers, given evidence that it does not perform as effectively in these cancers. (See 'Nonsquamous histology' above.)

As an example of available data, in a randomized trial in patients with advanced NSCLC, four platinum-containing regimens resulted in similar response rates and survival (cisplatin and paclitaxel, cisplatin and gemcitabine, cisplatin and docetaxel, or carboplatin and paclitaxel) [36]. Similarly, a separate trial suggests similar time to tumor progression and OS between vinorelbine and gemcitabine, when either is paired with cisplatin [37]. The incidence of phlebitis was higher in the vinorelbine-cisplatin arm, while thrombocytopenia, flu-like syndrome, and asthenia were more frequent in the gemcitabine-cisplatin arm.

PD-L1-HIGH TUMORS (AT LEAST 50 PERCENT)

Rationale for inclusion of immunotherapy — Randomized trials have demonstrated that a high level of programmed cell death ligand 1 (PD-L1) expression (at least 50 percent) predicts response to pembrolizumab, atezolizumab, or cemiplimab. This level of PD-L1 expression may be seen in approximately 30 percent of advanced NSCLCs [38]. As such, either immunotherapy alone or chemoimmunotherapy are options.

As observed in multiple trials discussed above, chemotherapy plus pembrolizumab improves progression free survival (PFS) relative to chemotherapy alone, with greatest benefit in those with PD-L1 high tumors. Survival benefits have also been observed, in some trials. (See 'Pembrolizumab plus chemotherapy (preferred)' above.)

Other data show survival benefits with checkpoint inhibitor monotherapy relative to chemotherapy in patients with PD-L1-high tumors below. (See 'Data regarding checkpoint inhibitor monotherapy' below.)

Choosing between checkpoint inhibitor monotherapy and chemoimmunotherapy — For patients with PD-L1-high rapidly progressive disease, or such a high tumor burden that early progression might lead to functional decline precluding chemotherapy in the second-line setting, we suggest concurrent use of a checkpoint inhibitor with chemotherapy. This approach has been associated with higher response rates than checkpoint inhibitor monotherapy. The choice of agents is analogous to those with PD-L1-unselected or low tumors. (See 'PD-L1 low (<50 percent) or unselected tumors' above.)

For those without rapidly progressive disease whose tumors have ≥50 percent PD-L1 expression, either checkpoint inhibitor monotherapy or chemoimmunotherapy are appropriate, and shared decision making is appropriate. In the absence of an overall survival (OS) benefit, those who value fewer burdens of treatment may choose immunotherapy monotherapy; patients who value delayed time to progression may opt for chemoimmunotherapy, as discussed above. (See 'PD-L1 low (<50 percent) or unselected tumors' above.)

Treatment with a checkpoint inhibitor plus chemotherapy has not been directly compared with checkpoint inhibitor monotherapy in those with PD-L1-high NSCLCs in a randomized trial. However, the following data are available:

●A meta-analysis of five randomized trials suggested, through indirect comparisons, that pembrolizumab plus chemotherapy was superior to pembrolizumab alone in regards to objective response rates (ORRs; relative risk 1.6, 95% CI 1.2-2.2) and PFS (hazard ratio [HR] 0.55, 95% CI 0.32-0.97) [39]. No statistically significant differences in OS were observed (HR 0.76, 95% CI 0.51-1.14).

●In a presented abstract of an analysis of 12 trials in patients with tumor PD-L1 ≥50 percent, half of them evaluating chemoimmunotherapy and the other half evaluating immunotherapy alone, median PFS was longer with chemoimmunotherapy than with immunotherapy (9.6 versus 7.1 months, HR 0.69, 95% CI 0.55-0.87) [40]. ORR was also better with chemoimmunotherapy (61 versus 43 percent). Although there was a trend towards improved OS with chemoimmunotherapy in the overall group (HR 0.82), improvements did not meet statistical significance, and among patients ≥75 years old there was a nonsignificant trend towards worsened survival with chemoimmunotherapy (HR 1.7).

Although some find these modest PFS benefits with incorporation of chemotherapy compelling, others see the lack of OS benefit a reason to favor immunotherapy alone.

Data regarding single agent immunotherapy versus single agent chemotherapy in patients who are not candidates for platinum-based chemotherapy are discussed below. (See 'Older or frail patients' below.)

Data regarding checkpoint inhibitor monotherapy — If checkpoint inhibitor monotherapy is chosen, either pembrolizumab, atezolizumab, or cemiplimab as single agents are appropriate [1]. Pembrolizumab has the practical advantage of regulatory approval for an every-six-week schedule based on data from the metastatic melanoma setting, as well as longer-term follow-up (discussed below). (See "Systemic treatment of metastatic melanoma lacking a BRAF mutation", section on 'Dosing considerations'.)

Nivolumab-plus-ipilimumab-based treatments are a reasonable alternative. (See 'Nivolumab plus ipilimumab, with or without chemotherapy' above and 'PD-L1 low (<50 percent) or unselected tumors' above.)

●Pembrolizumab – Pembrolizumab is approved by the US Food and Drug Administration (FDA) for the frontline treatment of patients with advanced epidermal growth factor receptor (EGFR)/anaplastic lymphoma kinase (ALK) wild-type NSCLC whose tumors have ≥50 percent PD-L1 expression based on the 22C3 pharmDx test [41].

Use of pembrolizumab for patients with PD-L1 expression ≥50 percent is supported by the phase III KEYNOTE-024 trial, in which 305 treatment-naïve patients with advanced NSCLC having at least 50 percent tumor cell PD-L1 staining were randomly assigned to pembrolizumab monotherapy (200 mg intravenously every three weeks) versus standard platinum-doublet chemotherapy [38]. At a median follow-up of 11.2 months, PFS, the primary endpoint, was prolonged with pembrolizumab compared with platinum-doublet chemotherapy (median PFS 10.3 versus 6 months; HR 0.50, 95% CI 0.37-0.68), and ORRs were improved (45 versus 28 percent). The median duration of response was not reached in the pembrolizumab group and was 6.3 months in the chemotherapy group. Severe (≥grade 3) treatment-related adverse effects were lower among patients receiving pembrolizumab (27 versus 53 percent). Any-grade pneumonitis was reported in 5.8 percent of patients treated with pembrolizumab, with severe pneumonitis in 2.6 percent.

Results at five-year follow-up demonstrate improved median OS with pembrolizumab (26.3 versus 13.4 months with chemotherapy; HR 0.62, 95% CI 0.48-0.81) [42].

Adding ipilimumab to pembrolizumab did not show improved efficacy relative to pembrolizumab monotherapy in a randomized trial in patients with high-PD-L1 NSCLC, but did increase toxicity [43]. As such, we do not use pembrolizumab in combination with ipilimumab; the combination of nivolumab plus ipilimumab is discussed above. (See 'Nivolumab plus ipilimumab, with or without chemotherapy' above.)

●Atezolizumab – Atezolizumab is FDA approved for the frontline treatment of those with PD-L1-stained ≥50 percent of tumor cells or PD-L1-stained tumor-infiltrating immune cells covering ≥10 percent of the tumor area, as determined by an FDA-approved test, with no EGFR or ALK genomic alterations [44].

This approval is based on results of IMpower 110 [44,45], which enrolled 572 treatment-naϊve patients with stage IV NSCLC expressing PD-L1 (PD-L1 stained ≥1 percent of tumor cells [TC] or tumor-infiltrating immune cells [IC] covering ≥1 percent of the tumor area). Among the subset of 205 patients with high PD-L1 (TC ≥50 percent or IC ≥10 percent), atezolizumab demonstrated improved OS relative to platinum-based chemotherapy (20 versus 13 months; HR 0.59, 95% CI 0.40-0.89). There were no statistically significant differences in OS for the other two PD-L1 subgroups evaluated (TC ≥5 percent or IC ≥5 percent; and TC ≥1 percent or IC ≥1 percent).

In the PD-L1-high subgroup, median PFS was 8.1 months with atezolizumab, versus 5 months with chemotherapy (HR 0.63, 95% CI 0.45-0.88); the confirmed ORR was 38 versus 29 percent, respectively.

Fatal adverse reactions occurred in 3.8 percent of the overall group of patients receiving atezolizumab, from causes including pulmonary embolism, acute myocardial infarction, and sepsis [45]. Grade ≥3 adverse events occurred in approximately 30 percent assigned to atezolizumab and 53 percent assigned to chemotherapy, with the most frequent in the atezolizumab group including pneumonia (2.4 percent), chronic obstructive pulmonary disease (1.7 percent), and pneumonitis (0.7 percent).

●Cemiplimab – Cemiplimab, a programmed cell death protein 1 inhibitor, improved OS and PFS compared with chemotherapy in patients with advanced NSCLC with PD-L1 of ≥50 percent without an EGFR, ROS1, or ALK genetic aberration [46], and has FDA approval for this indication [47].

In EMPOWER-Lung 1, among 565 patients with advanced NSCLC with PD-L1 of ≥50 percent, cemiplimab 350 mg every three weeks improved OS relative to platinum-doublet chemotherapy at 35 months follow-up (26 versus 13 months, respectively; HR 0.57, 95% CI 0.46-0.71), as well as PFS (8.1 versus 5.3 months; HR 0.51, 95% CI 0.42-0.62) [48]. OS and PFS were also improved in the intention-to-treat population of 712 patients (which additionally included patients with PD-L1 expression that had not been confirmed or had been established as <50 percent), despite 74 percent crossover [46]. Grade ≥3 toxicities were 28 percent with cemiplimab and 39 percent with chemotherapy.

DURATION OF THERAPY

Immunotherapy — In general, we continue treatment with a programmed cell death protein 1 (PD-1) axis inhibitor until progression or unacceptable toxicity occurs, although discontinuation after two years of treatment may be a reasonable alternative. For patients whose initial regimen includes platinum-based chemotherapy, duration of chemotherapy is discussed below (see 'Chemotherapy' below). The approach to duration of immunotherapy is based upon the randomized clinical trials leading to US Food and Drug Administration approval of PD-1 axis inhibitors in which respective agents were continued until progression.

Available data are limited but suggest that continuing checkpoint inhibitor therapy beyond one year is beneficial. The CheckMate 153 safety trial evaluated duration of nivolumab monotherapy, but only as an exploratory endpoint [49]. In this trial, patients with pretreated, advanced, PD-1 axis inhibitor-naïve NSCLC received nivolumab. Among 174 patients who completed a year of treatment without progression, those randomly assigned to continue nivolumab treatment versus observation experienced an improvement in progression-free survival (PFS; median 25 versus 9.4 months; hazard ratio [HR] 0.56, 95% CI 0.37-0.84), and overall survival (OS; not reached versus 32.5 months; HR 0.61, 95% CI 0.37-0.99).

However, other data suggest that discontinuation of pembrolizumab after two years may be an appropriate strategy [42,50,51]. As an example, in long-term follow-up of KEYNOTE-010, most patients who completed two years of pembrolizumab had a durable response (median three-year OS rate, 99 percent), with 64 percent of these patients demonstrating an ongoing response even after discontinuation of pembrolizumab [51]. Nearly one-half (43 percent) of patients who underwent "rechallenge" with pembrolizumab after discontinuation of therapy had a second response upon retreatment with pembrolizumab.

Similarly, in the five-year follow-up of KEYNOTE-024, in which patients were randomly assigned to chemotherapy or two years of pembrolizumab, five-year OS was doubled in those receiving pembrolizumab (32 versus 16 percent) [42]. A second course of pembrolizumab was allowed in this trial, for patients who completed two years of pembrolizumab or who stopped pembrolizumab after achieving a complete response and then had progression. Five-year follow-up of other trials also highlight a 40 to 44 percent rate of nonprogression and no further treatment at five years, among patients assigned to two years of treatment [5,8].

However, no trials have directly compared two years of immunotherapy versus ongoing treatment until progression. Additional clinical trials are needed, evaluating alternatives to indefinite continuous dosing, based upon the potential for durable responses to PD-1 axis inhibitors, continuous risk of toxicity, and the high cost of indefinite therapy.

Chemotherapy — For those in whom chemotherapy is indicated, our approach is as follows.

Four to six cycles of platinum doublet therapy — We administer the platinum doublet for four to six cycles.

Extending the duration of treatment with the initial platinum-based chemotherapy beyond four to six cycles is not recommended [26]. Although longer treatment duration increases PFS, it has at most only a moderate effect on OS [52].

A meta-analysis of five trials comparing six cycles versus three or four cycles failed to demonstrate a statistically significant improvement in OS (9.5 versus 8.7 months), although the difference in PFS was significant (6.1 versus 5.3 months) [53].

Maintenance therapy — For those receiving a platinum doublet in addition to immunotherapy, the checkpoint inhibitor maintenance is continued after the initial four to six cycles of chemotherapy. (See 'Immunotherapy' above.)

If tolerated, we also continue single-agent chemotherapy as maintenance, along with the checkpoint inhibitor, although many patients must discontinue at some point due to cumulative toxicity. Pemetrexed [54-62], docetaxel [60], and gemcitabine [61] all significantly prolong PFS when used as single-agent maintenance therapy after initial chemotherapy for advanced NSCLC; however, no data exist in regards to the benefit when paired with an immune checkpoint inhibitor.

Additionally, there are no randomized clinical trials that compare these three agents with each other as maintenance therapy. The most extensive data are from trials using pemetrexed, which we prefer for patients with nonsquamous carcinomas, particularly if pemetrexed was used in the frontline regimen. Docetaxel is an alternative to pemetrexed for patients with nonsquamous NSCLC. Either docetaxel or gemcitabine is preferred for patients with squamous cell carcinomas, a setting in which pemetrexed is not used.

●Is there a role for bevacizumab? – We do not incorporate bevacizumab maintenance, given lack of survival benefit and known risks.

Bevacizumab was previously sometimes included with a platinum-based doublet in the initial chemotherapy regimen for nonsquamous NSCLC. However, in a meta-analysis of four randomized controlled trials, addition of anti–vascular endothelial growth factor (VEGF) to pemetrexed maintenance failed to improve OS [63]. Per the American Society of Clinical Oncology living guideline, given the lack of a survival benefit and higher toxicities, anti-VEGF therapy should not be added to pemetrexed and platinum therapy or given as maintenance [64].

TOXICITY OF TREATMENT — Immune-related toxicities associated with programmed cell death protein 1, programmed cell death ligand 1, and cytotoxic T-lymphocyte antigen 4-blocking antibodies are discussed in more detail separately. (See "Toxicities associated with immune checkpoint inhibitors" and "Ocular side effects of systemically administered chemotherapy", section on 'Immune checkpoint inhibitors'.)

SPECIAL CONSIDERATIONS

Older or frail patients — Fit, older adult patients with good performance status have been found to benefit from systemic therapy for NSCLC to the same degree as younger patients. The general strategy of choosing a systemic agent for the fit, older patient is the same as for younger patients, and advanced age alone should not preclude receipt of therapy. However, the presence of comorbidities may influence choice of treatment. (See "Systemic chemotherapy for cancer in older adults" and "Systemic chemotherapy for cancer in older adults", section on 'Palliative chemotherapy'.)

Options for frail patients include supportive care, checkpoint inhibitors, or single-agent chemotherapy. Although pembrolizumab has an established role for patients with programmed cell death ligand 1 (PD-L1)-high tumors (see 'PD-L1-high tumors (at least 50 percent)' above), nivolumab plus ipilimumab has demonstrated a superior survival compared with chemotherapy, irrespective of PD-L1 expression, and may be an appropriate approach in patients for whom chemotherapy is not an option. (See 'Nivolumab plus ipilimumab, with or without chemotherapy' above.)

Finally, data suggest single-agent atezolizumab may be an appropriate alternative for those with PD-L1-low tumors and a poor performance status.

In the IPSOS trial, patients with advanced NSCLC ineligible for platinum-doublet containing chemotherapy (due to performance status ≥2 and/or comorbidities (table 4)), were assigned to first-line atezolizumab or single-agent chemotherapy (investigator's choice of gemcitabine or vinorelbine) [65]. Approximately 30 percent of these patients were 80 years or older, and 40 to 50 percent in each group had PD-L1 <1 percent. There was a modest benefit of atezolizumab over chemotherapy (median overall survival [OS] 10.3 versus 9.2 months, hazard ratio [HR] 0.78), comparable across all levels of tumor PD-L1 expression, and performance status (including those with performance status 3). Two-year OS rates were higher with atezolizumab than chemotherapy (24 versus 12 percent). Objective response rate also favored atezolizumab (17 versus 9 percent). Compared with chemotherapy, atezolizumab resulted in fewer grade 3 to 4 treatment-related adverse events (16 versus 33 percent) and treatment-related deaths (1 versus 3 percent). Although this strategy does not have regulatory approval, single-agent atezolizumab is a potential consideration for patients with negative tumor PD-L1 and a poor performance status.

Tools to assess the fitness of older adults for cancer treatment are discussed elsewhere. (See "Comprehensive geriatric assessment for patients with cancer" and "Systemic chemotherapy for cancer in older adults".)

Limitations in biomarkers — Although trials have demonstrated that tumor PD-L1 expression associates with increased likelihood of response to checkpoint inhibitors, it neither guarantees response in those with high tumor PD-L1 expression nor eliminates the possibility of response in those tumors that lack PD-L1 expression. There may be PD-L1 heterogeneity within tumors and between tumor sites and tumor PD-L1 expression can change over time in response to therapy. Nevertheless, based on the data from KEYNOTE-024 above, we recommend routine PD-L1 testing in all patients with newly diagnosed NSCLC to inform use of pembrolizumab monotherapy in the frontline setting.

Tumor mutational burden is another marker being evaluated in trials of checkpoint inhibitors, though its clinical utility has yet to be established.

Impact of steroids on efficacy of immunotherapy — Available data, largely from studies of immunotherapy in melanoma, suggest that corticosteroids for treatment of immune-related adverse events do not affect efficacy [66,67]. However, some evidence suggests that baseline corticosteroid use of ≥10 mg prednisone equivalent is associated with poorer outcomes with immunotherapy [68]. It remains unclear if the less favorable outcomes in these patients are directly related to the immunosuppressive effect of chronic corticosteroid use versus possibly a reflection of the unfavorable factors that necessitated use of steroids (eg, symptomatic brain metastases, weight loss, profound fatigue, or other factors) [69-71]. Although larger confirmatory studies are needed, we advise that corticosteroids be used sparingly and judiciously at the time of initiation of immunotherapy unless necessary for medical treatment (such as brain metastases).

In a retrospective review including 640 patients treated with single-agent programmed cell death protein 1 (PD-1)/PD-L1 blockade for advanced NSCLC, multivariate analysis suggested worsened OS (HR 1.3, 95% CI 1.03-1.57) and progression-free survival (PFS; HR 1.66, 95% CI 1.3-2.2) among the 14 percent of patients who received corticosteroids of ≥10 mg of prednisone equivalent daily [68]. Similarly, in a separate retrospective review, both median PFS and OS were shorter among patients receiving ≥10 mg of prednisone at the time of immunotherapy initiation relative to those receiving lesser amounts or no steroids [69]. However, when analyzed according to the reason for steroid administration, the worsened survival outcomes were shorter only for patients receiving ≥10 mg of prednisone for palliative indications rather than for cancer-unrelated reasons.

Overall, these data suggest that the worsened outcomes observed in patients receiving steroids with immunotherapy may not be secondary to the steroids themselves, but rather are driven by other poor prognostic factors. Further data are needed.

Patterns of response and progression — The patterns of response to treatment with immunotherapy agents can differ from those with molecularly targeted agents or cytotoxic chemotherapy. For example, patients may have a transient worsening of radiologic lesions prior to disease stabilization or regression, though this is an uncommon presentation, and any evidence of clinical worsening makes this unlikely to be the case. This is discussed in more detail elsewhere. (See "Principles of cancer immunotherapy", section on 'Patterns of response'.)

Additionally, hyperprogressive disease, in which an acceleration of tumor growth occurs upon initiation of a given therapy, may occur among patients treated with immunotherapy, though this concept remains the subject of some debate even among experts in immunotherapy. If hyperprogressive disease is suspected, for example if tumor burden doubles or more at the first restaging scan, prompt recognition and transition to chemotherapy is warranted. In a retrospective study of over 400 previously treated patients with advanced NSCLC, hyperprogressive disease, defined as tumor growth rate exceeding 50 percent per month, occurred in 14 percent of those treated with PD-1/PD-L1 inhibitors and 5 percent of those treated with single-agent chemotherapy [72].

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Diagnosis and management of lung cancer".)

SUMMARY AND RECOMMENDATIONS

●Factors in choosing initial therapy – Lung cancer is the leading cause of cancer-related mortality in the United States and worldwide. More than 80 percent of lung cancers are classified as non-small cell lung cancer (NSCLC). Immune checkpoint inhibitors targeting either programmed cell death protein 1 (PD-1) or programmed cell death ligand 1 (PD-L1) have become integrated into the clinical approach for management of NSCLC. (See 'Factors in choosing initial therapy' above.)

●Molecular assessment – For patients who have not received systemic therapy for advanced NSCLC, we assess tumor PD-L1 as well as evaluate for targetable driver mutations. The management of NSCLCs associated with targetable driver mutations is discussed elsewhere. (See "Anaplastic lymphoma kinase (ALK) fusion oncogene positive non-small cell lung cancer" and "Systemic therapy for advanced non-small cell lung cancer with an activating mutation in the epidermal growth factor receptor" and "Personalized, genotype-directed therapy for advanced non-small cell lung cancer".)

●Initial treatment for PD-L1-unselected tumors, or with PD-L1 <50 percent expression – For those without a targetable driver alteration and either <50 percent of tumor cells staining for PD-L1 or unknown PD-L1 expression, we recommend platinum-based chemotherapy combined with an immune checkpoint inhibitor, rather than chemotherapy alone (Grade 1B). For choice of immune checkpoint inhibitor, we typically suggest pembrolizumab (table 5) (Grade 2C), given that it has regulatory approval for an every-six-week schedule, as well as longer-term follow-up; however we note that other immune checkpoint inhibitors with chemotherapy, or nivolumab plus ipilimumab as a nonchemotherapy option, are reasonable alternatives.

●Initial treatment for PD-L1 ≥50 percent expression – For patients without a targetable driver alteration in whom at least 50 percent of tumor cells stain for PD-L1, we recommend a checkpoint inhibitor, with or without chemotherapy, rather than chemotherapy alone (Grade 1A). For choice of immune checkpoint inhibitor, we typically suggest pembrolizumab (table 5) (Grade 2C), given that it has regulatory approval for an every-six-week schedule, as well as longer-term follow-up; however we note that other immune checkpoint inhibitors with supporting randomized trial data are also reasonable options.

•The decision regarding when to add chemotherapy to PD-1/PD-L1 blockade for those with PD-L1-high tumors depends upon the burden of disease and pace of progression. For those with rapidly progressive disease, or such a high tumor burden that early progression might lead to functional decline precluding second-line chemotherapy, we suggest the addition of concurrent chemotherapy to an immune checkpoint inhibitor (Grade 2C). For those without rapidly progressive disease, either checkpoint inhibitor monotherapy or chemoimmunotherapy are appropriate. (See 'PD-L1-high tumors (at least 50 percent)' above.)

●Choice of chemotherapy – For patients whose initial treatment strategy includes chemotherapy, we suggest using a carboplatin-based doublet rather than other regimens (Grade 2C). However, a cisplatin-based doublet is a reasonable alternative in appropriate patients. Multiple agents pair effectively with platinums. For those with nonsquamous cancers, we suggest the addition of pemetrexed to the platinum agent (Grade 2C). For squamous cancers, options for the second agent include taxanes (docetaxel, paclitaxel, nanoparticle albumin-bound paclitaxel [nabpaclitaxel]), gemcitabine, vinorelbine, and camptothecins (irinotecan, topotecan). (See 'Choice of chemotherapy to pair with checkpoint inhibitor' above.)

●Duration of treatment

•For patients treated with a PD-1 axis inhibitor, we suggest continued treatment until progression or unacceptable toxicity occurs (Grade 2C). Alternatively, one may reasonably discontinue after two years, although this has not been prospectively compared with indefinite treatment. (See 'Immunotherapy' above.)

•For those receiving a platinum doublet, it is continued for four to six cycles.

-We also suggest continuation of single-agent nonplatinum chemotherapy as maintenance (Grade 2C), alongside the immune checkpoint inhibitor, although we acknowledge that data are lacking regarding the benefit of maintenance chemotherapy and many patients discontinue due to cumulative toxicity. (See 'Maintenance therapy' above.)

ACKNOWLEDGMENTS — The UpToDate editorial staff acknowledges Scott Gettinger, MD, Matthew Hellmann, MD, and Howard (Jack) West, MD, who contributed to earlier versions of this topic review.