Methodology

The unbiased and reliable OS endpoint was used throughout the present investigation. There is growing recognition among the oncology community that OS gains by anticancer drugs of < 2 months might not be clinically significant. In our study, grade D was assigned to < 60 days and A+ to > 240 days. Our goal was to facilitate full disclosure and transparency of outcome, costs and value to patients [20, 21]. The proposed methodology was not designed for medical economists but was rather geared towards the community oncologists. The simplicity of the grading system could enable oncologists, pharmacists and nurses to rapidly calculate drug value. The value was expressed relative to $100,000, the average acceptable ACER in the US. Admittedly, there are drawbacks limiting applications of the proposed methodology. Costs of hospitalization, drug administration, long-term treatment of chronic AEs and global QoL measures were not included. Fees for professionals including physicians, pharmacists and nurses were not considered. The methodology also failed to capture and capitalize on the longer duration of response of the ICPI over Doc.

Previously published methodology measuring differences in cost vs. outcome between 2 interventions [22, 23] could not be applied since in the present investigation 5 drugs were compared. Generic Doc costs were, not surprisingly much cheaper as compared with the ICPI. There was a non-significant cost difference between Nivo, Atezo and Pembro. The 4-week costs of Doc were $306, Nivo 240 mg $11,914, Atezo 1200 mg $11,493 and Pembro 200 mg $11,509. The incremental increase in cost per one-day OS gain was therefore used as the basis of our calculations. Costs, value, hazard ratios (HR) and median OS of Nivo at 12 months-milestone have recently been reported in a wide variety of cancer [15].

AEs

Docetaxel demonstrated AEs gr 3/4 of > 20%. The average AEsTC of 6 cycles were $1978 and 12 cycles $4940. The AEsTC could be burdensome and costly [24]. The cascade of neutropenia to febrile neutropenia, the non-hematological AEs and peripheral neuropathy have been well documented [1].

The gr ¾ AEs of the ICPI were generally < 10% [4–10]. The oral and iv steroids used for treatment of AEs were inexpensive. The costs of treatment represented an insignificant fraction of the total drug costs. With early use of scans and timely interventions, the average yearly AEsTC per patient treated by the ICPI were estimated at $1480. In CheckMate 017 in the squamous-NSCLC, the AEsTC of Nivo were $439 vs. docetaxel of $7024. In CheckMate 057 in non-squamous, Nivo AEsTC were $518, vs. Doc $5940 [25]. There has been wide spread awareness of the potential hazards with continued gain in experience and management of AEs since the introduction of bevacizumab [26], Doc and the ICPI.

In our study, Nivo demonstrated discernible value improvement of the squamous over the non-squamous histology. Atezolizumab and Pembro seemed to be equally effective irrespective of histology. Of interest, there is no documentation yet of benefit by PD-L1 enrichment in the squamous type. The latter is generally considered more homogenous with a higher incidence in non-smokers and males than the non-squamous.

Cost vs. value

The balance between drug costs and value is sensitive but elusive. The debate seems to continue [27]. Costs need to be maintained at certain levels for drug companies to proposer and keep their innovative edge. It is estimated that 1 billion dollars is spent to successfully bring a formula from the drawing boards to pharmacy shelves. Nonetheless, the price tags of 10 mg/Kg of Nivo and Pembro could dampen their use. The affordability of the ICPI in combination with other immune drugs e.g. Ipilimumab would be negatively impacted and curtailed using relatively high doses and frequent administrations. Such costs could be only tolerated if long-term disease control and/or cure are achieved.

The pendulum however has recently tipped in favor of value since ASCO and ESMO issued their initiatives in 2013 [17, 18]. The National Institute for Health and Care Excellence (NICE) in the United Kingdom stipulates that ACER interventions ought not to exceed 20,000–30,000 pounds per quality-adjusted life-year gained. Of note, Nivo approval by NICE is pending and negotiations are still continuing. It would be helpful to oncology professionals, patients, drug authorities and society if the FDA could give further guidance on value. At present, the US law prohibits discussion of drug costs during the approval process.

PD-L1 biomarker and drug value

In the early clinical studies of the ICPI in NSCLC various PD-L1 expression in tumor cells and infiltrating immune cells. At present, PD-L1 levels are being standardized and levels expressed as tumor proportion scores (TPS). There is up to 30% response in patients with PD-L1 enhanced expression. The findings of durable responses in non-enriched PDL-1 are indicative of the “imperfections” of PDL-1 marker. The tumor mutation load in combination with PDL-1, are emerging as more precise predictive markers [28].

The present study was the first to demonstrate that PDL-1 enrichment resulted in improved OS and enhanced value in NSCLC. The OS increased from < 100 days with C and D grade to > 150–300 days with B and A+. The OS increase by PDL-1 enrichment was unexpected and unprecedented in an incurable disease as NSCLC. From an economic point of view, it would be prudent to spend upfront few hundred dollars on a reliable predictive marker or set of markers than to bear the costs of treating “all comers”. This strategy would spare the non- responders the high costs and unnecessary toxicity of ineffective therapy.

Duration of treatment

The ICPI treatment paradigm is evolving. The current recommendations on use of the ICPI are to continue treatment till AEs occurrence, progression or beyond progression. Costs and to a lesser extent AEs would increase with extended use. The duration of treatment comes down to a choice between cost vs. value and is presently an area of active investigation. It remains to be seen whether the incremental survival benefit by prolonged use ICPI could justify the additional yearly- costs of $131,000.

Policies and perspectives

The FDA has recently adopted a new trend of approval of new drugs based on early endpoints other than survival. The present methodology could be modified to use disease control rates and other endpoints after adjustments with appropriate correction factors. Higher doses and longer duration of treatment by costly drugs need to be scrutinized before utilization.

In 2nd-line non- mutated NSCLC, the superiority of ICPI over chemotherapy has been generally accepted despite concerns on cost issues. Costs of Nivo and Pembro were described by Salts [29] as thousands of times the cost of gold.

In 1st-line, Pembro in combination with pemetrexed and carboplatin without PD-L1 enrichment has recently demonstrated higher response rates with deeper and longer duration of response [30]. The overall survival data have not yet matured. Clearly, PD-L1 is not the whole story. Tumor mutation burden can be used as well to decide on immunotherapy approaches [31]. Enthusiastic use of the ICPI however needs to be tempered by the economic pressures of countries and patients with limited resources. With the upcoming waves of generic chemotherapeutic drugs including pemetrexed, the role of Doc and many others is not yet dead but just diminished [32].